SVD identifies transcript length distribution functions from DNA microarray data and reveals evolutionary forces...

Area of Science:

• Molecular Biology
• Evolutionary Biology
• Bioinformatics

Background:

• Transcript length is a fundamental characteristic of genes.
• Understanding the evolutionary pressures on transcript length is crucial for deciphering gene regulation.
• Previous hypotheses suggested evolutionary forces, akin to a harmonic oscillator, influence transcript length.

Purpose of the Study:

• To identify evolutionary forces acting on transcript length.
• To investigate the relationship between transcript length, gene ontology, and cellular metabolism.
• To explore transcript length regulation in normal versus glioblastoma multiforme (GBM) tissues.

Main Methods:

• Singular Value Decomposition (SVD) applied to DNA microarray data to determine transcript length distribution functions.
• Comparison of transcript length distributions across gene ontology annotations in human and yeast.
• Analysis of transcript length differences in GBM versus normal brain tissue using The Cancer Genome Atlas data.

Main Results:

• SVD successfully identified transcript length distribution functions as "asymmetric generalized coherent states" without prior assumptions.
• Transcripts involved in protein synthesis and mitochondrial metabolism are significantly shorter than those in glucose metabolism across human and yeast.
• Glioblastoma multiforme (GBM) exhibits altered transcript length regulation, favoring shorter transcripts for protein synthesis/mitochondrial metabolism and suppressing longer ones for glucose metabolism/brain activity.

Conclusions:

• Transcript length is significantly correlated with cellular functions like metabolism and gene ontology.
• Differential regulation of transcript length may represent a physical mechanism for metabolic control in normal and tumor cells.
• Findings support evolutionary models proposing harmonic oscillator-like forces shaping transcript length.